Movement Lab Evidence Report

Generated 2026-07-14T11:25:27.075Z. Canonical model: pro_prototype.

QA scenarios10/10
Unit tests7
Parity fixtures7
Parity gatePASS
Trajectory laboratory only — no OS pointer injection, no browser event automation. L4–L5 parity with Python is structural (RNG differs: mulberry32 vs numpy PCG64).

Session scenario (Precursor-style journey)

Multi-target sign-in session · seed=7 · captured 2026-07-14T11:44:21.461Z

Session L0
L0 session replay (GIF)
Session L5
L5 session replay (GIF)
LevelDuration (ms)Path length (px)Speed peaksMean reaction (ms)Mean dwell (ms)
L0 11444 1549 0 237 574
L5 9090 2078 15 283 574

Session path length by level (seed=7)

LevelTotal path (px)Duration (ms)
L0154911444
L115498742
L215498742
L315508742
L414869090
L520789090

Run npm run evidence:session to regenerate local GIF artifacts (gitignored).

Prototype decision

pro_prototype selected (95/100 weighted score). See evidence/decision.md.

QA scenarios S01–S10

IDActionResultScreenshot
S01 Load lab, default L5 PASS S01
S02 Switch L0→L5 presets PASS —
S03 Shift-click start, click target PASS S03
S04 Toggle feature switches PASS —
S05 Animate trajectory PASS S05
S06 Overlay L0–L5 PASS S06
S07 Export CSV PASS —
S08 Export JSON config PASS —
S09 Fixed seed repeat PASS —
S10 Narrow vs wide target duration PASS —

Metrics by level (seed=7)

LevelDuration (ms)Path efficiencyPeak speed (px/s)Norm. jerk
L0 620.0 1.0000 2300.3 21730137.00
L1 413.5 1.0000 3449.0 6703125.00
L2 413.5 1.0000 3230.0 1475.99
L3 413.5 0.9992 5305.1 12967.77
L4 413.5 0.9980 5461.4 13159.43
L5 413.5 0.8959 6609.5 24955.81

Level screenshots

Level L0
L0
Level L1
L1
Level L2
L2
Level L3
L3
Level L4
L4
Level L5
L5

Overlay

L0-L5 overlay